This dataset mirrored from http://18.104.22.168/WebUI/CatDatabase/catData.html, which circa May 2017 is a dead link. The original page is available in Wayback: https://web.archive.org/web/20150520175645/http://22.214.171.124/WebUI/CatDatabase/catData.html
The CAT dataset includes 10,000 cat images. For each image, we annotate the head of cat with nine points, two for eyes, one for mouth, and six for ears. The detail configuration of the annotation was shown in Figure 6 of the original paper:
Weiwei Zhang, Jian Sun, and Xiaoou Tang, ＾Cat Head Detection - How to Effectively Exploit Shape and Texture Features￣, Proc. of European Conf. Computer Vision, vol. 4, pp.802-816, 2008.
The annotation data are stored in a file with the name of the corresponding cat image plus ＾.cat￣, one annotation file for each cat image. For each annotation file, the annotation data are stored in the following sequence ！ (Number of points, default is 9), (Left Eye), (Right Eye), (Mouth), (Left Ear-1), (Left Ear-2), (Left Ear-3), (Right Ear-1), (Right Ear-2), and (Right Ear-3).
Training, Validation, and Testing
We randomly divide the data into three sets: 5,000 images for training, 2,000 images for validation and 3000 images for testing.
The CAT dataset is only for research purposes, we do not have any copyright of the images.
"Error correction: replace CAT_00/00000003_019.jpg.cat with the file: 00000003_015.jpg.cat . (Thanks to Jiayong from Google who found the error.)"